Project health degree assessment method and system, electronic equipment and medium

Abstract

The invention provides a project health degree assessment method and system, electronic equipment and a storage medium, and aims to solve the problems of single project assessment dimension, low accuracy and poor adaptability. The method comprises the steps of selecting a project needing to be monitored according to an actual business scene, and constructing a model data set; the gross profit rate condition, the account receivable and the milestone performance information of the project are integrated, and each subitem business scoring index corresponding to the gross profit rate condition, the account receivable and the milestone performance dimension is constructed; fitting a dynamic relationship between a business scoring index and a score through a nonlinear function, and optimizing a fitting parameter by using an actual model data set so as to convert the business scoring index into a standardized health degree score; determining a subitem weight of each business scoring index; and calculating the health degree score of the project according to the scoring function and the determined subitem weight by adopting the latest data up to the current account period. According to the invention, the health state of the project can be accurately judged, and the scientificity and refinement level of project management can be improved.

Description

Technical Field

[0001] The present disclosure relates to the technical field of data processing and evaluation, and particularly to a method for evaluating project health, a system for evaluating project health, an electronic device, and a computer-readable storage medium. Background Art

[0002] The evaluation of the health of operator ICT projects is a systematic project management tool, aiming to comprehensively diagnose the operating status of information and communication technology (ICT) projects, identify potential risks, evaluate whether the project is progressing as planned, whether resources are reasonably allocated, and whether it can continuously create value, and ultimately ensure that the project is consistent with the operator's strategic goals. The evaluation of project health is the core link connecting project execution and strategic implementation; it can help operators quickly respond to changes in a complex environment, ensure that the project achieves the expected goals and continuously creates value.

[0003] However, there are still some problems in the existing project health evaluation, such as single evaluation dimension: the existing technology often evaluates the project health degree based on only a limited number of dimensions (such as accounts receivable, gross profit margin, etc.), which leads to one-sided evaluation results and cannot comprehensively reflect the true project health situation; low accuracy: when calculating the index values, the existing technology often fails to capture the change law of the corresponding health degree when the business changes, resulting in limited accuracy and reliability of the model; poor adaptability: the existing health evaluation models usually have difficulty adapting to the changes in different scenarios and requirements, lacking flexibility and scalability. Therefore, there is currently a lack of a more scientific and objective method for evaluating the health of operator ICT projects. Summary of the Invention

[0004] In order to at least solve the problems of single evaluation dimension, low accuracy, and poor adaptability of the existing project health evaluation methods. The present disclosure provides a method for evaluating project health, a system for evaluating project health, an electronic device, and a computer-readable storage medium; by evaluating the project health degree through multiple dimensions, and using the non-linear function method to more accurately depict the impact of index changes on the health degree, the scoring model can be made more adaptable to business changes, accurately judge the current health status of the project, and improve the scientific and refined level of project management.

[0005] In a first aspect, the present disclosure provides a method for evaluating project health, the method comprising:

[0006] Select projects to be monitored according to the actual business scenario, and construct a model data set;

[0007] Based on the gross profit margin status, accounts receivable, and milestone fulfillment information of comprehensive projects, construct various sub - item business scoring indicators corresponding to the gross profit margin status dimension, accounts receivable dimension, and milestone fulfillment dimension respectively;

[0008] Fit the dynamic relationship between business scoring indicators and scores through non - linear functions, and optimize the fitting parameters using actual model data sets to convert business scoring indicators into standardized health scores, making the scores more in line with business rules;

[0009] Determine the sub - item weights of each business scoring indicator;

[0010] Adopt the latest data as of the current accounting period and calculate the project health score according to the scoring function and the determined sub - item weights.

[0011] Furthermore, the selection of projects to be monitored according to the actual business scenario and the construction of the model data set include:

[0012] When selecting projects to be monitored, eliminate projects whose establishment exceeds the preset time and projects in abnormal states;

[0013] Obtain the original data of projects from the data center in the form of subscribing to data sets;

[0014] Obtain the basic project information, milestone information, financial information, and information related to the age of accounts receivable of each project from the original data.

[0015] Furthermore, the method also includes:

[0016] Conduct exploratory data analysis on the obtained original data, including:

[0017] Conduct data checks, including: consistency with the data dictionary, sufficiency of information, data understanding, data availability, data relevance;

[0018] Accounts receivable analysis: Analyze the scale of accounts receivable, the changing trend of the aging structure from the aspects of province, project type, and statistical month;

[0019] Based on statistical reports, view the increasing and decreasing trend and interval distribution of revenue;

[0020] Based on statistical reports, view the distribution of gross profit margin and its changing trend in different years;

[0021] Based on statistical reports, view the overall proportion of delayed collection, increasing and decreasing trend, and days distribution.

[0022] Furthermore, the business scoring indicators include:

[0023] The established gross profit margin and the executed gross profit margin in the gross profit margin status dimension;

[0024] The proportion of accounts receivable to revenue and the aging structure in the accounts receivable dimension;

[0025] In the milestone fulfillment dimension, the current number of days of delay calculation, the number of milestones with delayed collection, the maximum value of historical days of delayed collection, and the historical number of milestones with delayed collection.

[0026] Furthermore, the dynamic relationship between the business scoring metrics and the scores is fitted by a non-linear function, and the fitting parameters are optimized using the actual model data set, including:

[0027] Establish the upper and lower bounds of the metrics: Use the data within the observation period to analyze the distribution of each metric in the data, and combine with the actual business situation to determine the upper and lower bounds of the metric score changes. The metric scores outside the upper and lower bounds are given full marks or zero marks;

[0028] Prepare the data corresponding to the scores: Combine the distribution of metrics related to health and expert opinions to construct a data set;

[0029] Define the non-linear model: According to the law of the health degree of the business changing with the metrics, determine the corresponding non-linear function and concavity / convexity, and define the MODEL function;

[0030] Fit the model: For each metric, perform model fitting through the corresponding non-linear function;

[0031] Evaluate the model: Visualize the fitting effect of the non-linear scoring function by plotting the actual data points and the fitting curve;

[0032] Determine the parameters: Judge the fitting effect through the visualized data and the fitting curve, and determine the final parameter values;

[0033] Calculate the aging structure coefficient: Establish the historical rolling rate data of the aging, and construct an MA model to calculate the coefficients for different aging periods.

[0034] Furthermore, the determination of the sub-item weights of each business scoring metric includes:

[0035] Generate the appropriate weight interval for each metric according to expert evaluation and AHP (Analytic Hierarchy Process), including:

[0036] Construct a judgment matrix by having experts compare the importance of each metric pairwise;

[0037] Calculate the weights given by each expert through the geometric mean method and perform normalization processing. Based on the mean and standard deviation of the weights, and combined with relevant rule-based business adjustments, set the corresponding weight constraint intervals;

[0038] Perform weight optimization based on the Lagrange multiplier method to find the optimal ratio of the weights of the multi-dimensional scoring system, including:

[0039] Construct a project scoring objective function based on multiple indicators, and define the normalization constraints and interval constraints for each weight simultaneously;

[0040] Construct the Lagrangian function and introduce the Lagrangian multiplier system, and through the strict mathematical framework of the KKT conditions (Karush-Kuhn-Tucker Conditions), realize the modeling and solution of the multi-constraint optimization problem, ensuring that the optimization result not only meets the extreme value requirements of the objective function, but also conforms to the preset weight constraint conditions.

[0041] Further, the method further includes:

[0042] Determine the health level of the project according to the measurement result of the project's health score with reference to the project score grading standard;

[0043] Warn about projects with a significant decline in scores or projects that do not meet the health level standard.

[0044] In a second aspect, the present disclosure provides a project health assessment system, and the system includes:

[0045] A data selection module, which is configured to select projects to be monitored according to the actual business scenario and construct a model data set;

[0046] A scoring index construction module, which is configured to comprehensively consider the gross profit margin status, accounts receivable, and milestone fulfillment information of the project, and construct each sub-item business scoring index corresponding to the gross profit margin status dimension, accounts receivable dimension, and milestone fulfillment dimension;

[0047] A scoring function fitting module, which is configured to fit the dynamic relationship between the business scoring index and the score through a non-linear function, and optimize the fitting parameters using the actual model data set, so as to convert the business scoring index into a standardized health score, making the score more in line with business rules;

[0048] A sub-item weight determination module, which is configured to determine the sub-item weights of each business scoring index;

[0049] A health measurement module, which is configured to use the latest data as of the current accounting period and calculate the health score of the project based on the scoring function and the determined sub-item weights.

[0050] In a third aspect, the present disclosure provides an electronic device, including a memory and a processor, where a computer program is stored in the memory, and when the processor runs the computer program stored in the memory, the processor executes the project health assessment method according to any one of the first aspects.

[0051] Fourthly, the present disclosure provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the item health assessment method described in any one of the above first aspects is implemented.

[0052] Beneficial effects:

[0053] The item health assessment method, item health assessment system, electronic device and storage medium provided by the present disclosure; through multi-dimensional data fusion and dynamic weight optimization, the comprehensiveness and accuracy of the assessment are improved. Through the non-linear function method, the impact of index changes on the health degree is more accurately characterized. Combining expert experience with mathematical optimization methods ensures the professionalism and scientificity of the assessment index setting, enhances the adaptability, accuracy and reliability of the model to business changes; can make the scoring model more adaptable to business changes, conform to the actual business logic, accurately judge the current health status of the project, and improve the scientific and refined level of project management. Description of the drawings

[0054] Figure 1 It is a schematic flowchart of an item health assessment method provided by Embodiment 1 of the present disclosure;

[0055] Figure 2 It is a sample data diagram of milestone information data provided by an embodiment of the present disclosure;

[0056] Figure 3 It is a schematic diagram of a scoring function and coefficient measurement sample provided by an embodiment of the present disclosure;

[0057] Figure 4 It is a schematic diagram of the corresponding weight constraint interval of each dimension scoring index provided by an embodiment of the present disclosure;

[0058] Figure 5 It is a schematic diagram of the weight distribution result of a scoring index provided by an embodiment of the present disclosure;

[0059] Figure 6 It is a schematic diagram of a project score grading standard provided by an embodiment of the present disclosure;

[0060] Figure 7 It is a schematic diagram of an overall scheme process provided by an embodiment of the present disclosure;

[0061] Figure 8 It is a scoring sample diagram of some projects in a certain area provided by an embodiment of the present disclosure;

[0062] Figure 9 It is an architecture diagram of an item health assessment system provided by Embodiment 2 of the present disclosure;

[0063] Figure 10It is an architecture diagram of an electronic device provided in Embodiment 3 of the present disclosure. Detailed implementation manners

[0064] To enable those skilled in the art to better understand the technical solutions of the present disclosure, the present disclosure will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments and accompanying drawings described herein are only for explaining the present invention, rather than limiting the present invention.

[0065] It should be noted that the terms "first", "second", etc. in the specification and claims of the present disclosure and the above-mentioned accompanying drawings are used to distinguish similar objects, and do not necessarily need to describe a specific order or sequence; and, without conflict, the embodiments in the present disclosure and the features in the embodiments can be combined with each other arbitrarily.

[0066] Among them, the terms used in the embodiments of the present disclosure are only for the purpose of describing specific embodiments, and are not intended to limit the present disclosure. The singular forms "a", "the", and "said" used in the embodiments of the present disclosure and the appended claims are also intended to include the plural forms, unless the context clearly indicates otherwise.

[0067] In the subsequent description, the suffixes such as "module", "component", or "unit" used to represent elements are only for the convenience of the description of the present disclosure, and have no specific meaning in itself. Therefore, "module", "component", or "unit" can be used interchangeably.

[0068] The technical solutions of the present disclosure and how the technical solutions of the present disclosure solve the technical problems existing in the prior art will be described in detail below with specific embodiments. It can be understood that in the embodiments of the present application, the execution subject may execute some or all of the steps in the embodiments of the present application. These steps or operations are only examples, and the embodiments of the present application may also execute other operations or various deformations of the operations. In addition, the various steps may be executed in different orders presented in the embodiments of the present application, and it is possible that not all of the operations in the embodiments of the present application need to be executed. And, these several specific embodiments below can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments.

[0069] Figure 1 It is a schematic flowchart of a project health assessment method provided in Embodiment 1 of the present disclosure. As Figure 1 shown, the method includes:

[0070] Step S101: Select items to be monitored according to the actual business scenario, and construct a model data set;

[0071] Step S102: Construct respective sub - item business scoring indicators corresponding to the gross profit margin dimension, accounts receivable dimension, and milestone fulfillment dimension by integrating the gross profit margin status, accounts receivable, and milestone fulfillment information of the project.

[0072] Step S103: Fit the dynamic relationship between the business scoring indicators and the scores through a non - linear function, and optimize the fitting parameters using the actual model data set to convert the business scoring indicators into standardized health scores, making the scores more in line with business rules.

[0073] Step S104: Determine the sub - item weights of each business scoring indicator.

[0074] Step S105: Use the latest data as of the current accounting period to calculate the project health score according to the scoring function and the determined sub - item weights.

[0075] The purpose of the embodiments of the present disclosure is to construct a project health assessment model. Through business scoring indicators in multiple dimensions such as gross profit margin status, accounts receivable, and milestone fulfillment, using the non - linear function method and the optimal order, it more accurately depicts the impact of indicator changes on the health level, and dynamically determines the sub - item weights of each business scoring indicator through expert experience and data rules, providing a training basis for the health assessment model, thereby accurately judging the current health status of the project and providing core empowerment for management scenarios such as project risk control and accounts receivable control. The embodiments of the present disclosure can be applied to the assessment of various projects, especially ICT projects.

[0076] Taking the ICT project as an example, to achieve the above - mentioned purpose, the embodiments of the present disclosure first need to perform sample selection, select the projects to be monitored according to the actual business scenario, construct a model data set, so as to screen out effective project data and perform data pre - processing.

[0077] To more comprehensively evaluate the project health, construct an ICT project health evaluation model from dimensions such as project gross profit margin, milestone fulfillment, and accounts receivable, and construct corresponding sub - item business scoring indicators (referred to as scoring indicators) for each dimension. Through scoring indicators covering multiple dimensions such as finance, progress, and risk, achieve an all - round scan of the ICT project health and comprehensively reflect the true project health.

[0078] Different from the traditional interval fixed-point method and linear function method, the embodiments of the present disclosure use the non-linear function method to fit the dynamic relationship between business scoring indicators and scores, and optimize the fitting parameters using the actual model data set. The actual model data set is obtained from the real project data in sample selection. Combining expert experience to score the indicators to form actual data points to test the fitting curve, and using the real data points as the validation set to evaluate the fitting effect, which can ensure the accuracy and reliability of the model in the actual scenario; adopting the optimal order to more accurately depict the impact of indicator changes on the health level. Then determine the sub-item weights of each business scoring indicator according to the actual situation. Through the non-linear scoring function, the business scoring indicators are converted into standardized health scores, ensuring that the scoring function not only conforms to the data law but also is close to the business reality, and improving the reliability of the model by optimizing the weight allocation, so as to realize a more accurate project health evaluation.

[0079] For each project to be evaluated (including the projects that need to be monitored and are in an effective state in the current accounting period and new projects), the latest data as of the current accounting period is adopted. In the model, input the basic information of the project, milestone information, financial information such as revenue and cost, and data related to the accounts receivable aging; calculate the project health score according to the scoring function and the determined sub-item weights. According to the obtained scoring results, the health status of the project can be determined. The higher the score, the better its health level, and its health level is divided according to the score range.

[0080] For the health status of each obtained project, through a visualization display tool, such as developing various types of dashboards (such as large-screen display, ICT view management module) to intuitively present the health level of the projects within the jurisdiction of each region, which is convenient for the management to effectively manage the overall risk.

[0081] The embodiments of the present disclosure improve the comprehensiveness and accuracy of the evaluation through multi-dimensional data fusion and dynamic weight optimization. Through the non-linear function method, it more accurately depicts the impact of indicator changes on the health level. Combining expert experience and mathematical optimization methods, it ensures the professionalism and scientificity of the evaluation indicator setting, enhances the adaptability, accuracy and reliability of the model to business changes; can make the scoring model more adaptable to business changes, conform to the actual business logic, accurately judge the current health status of the project, and improve the scientific and refined level of project management. Through systematic health evaluation, operators can dynamically optimize project management, reduce the failure risk, and maximize the commercial value and social benefits.

[0082] Further, the selection of the projects that need to be monitored according to the actual business scenario and the construction of the model data set include:

[0083] When selecting the projects that need to be monitored, eliminate the projects whose establishment time exceeds the preset time and the projects in an abnormal state;

[0084] Obtain the original data of the project from the data middle platform in the form of subscribing to data sets;

[0085] Obtain the basic project information, milestone information, financial information, and information related to the aging of accounts receivable for each project from the original data.

[0086] Projects that exceed the preset time (such as more than 5 years since project establishment) are excluded from the modeling samples, and projects with abnormal statuses such as "project approval in progress" and "project terminated" are also excluded.

[0087] Projects that exceed the preset time may affect the model generalization ability due to changes in the business environment or loss of data validity; projects in the process of project approval have not entered the execution stage, have no actual progress and financial data, and cannot be evaluated for health; projects that have been terminated have incomplete data and may contain extreme values; by excluding these projects, noise data interference can be avoided and the model accuracy can be improved;

[0088] The data middle platform integrates data sources, including business system and external system data, and extracts data regularly in the form of subscribing to data sets, such as through API (Application Programming Interface) or ETL (Extract-Transform-Load) tools, to ensure data real-time.

[0089] After data exclusion, the basic project information, milestone information, financial information, and information related to the aging of accounts receivable for each project can be obtained through data subscription, so as to ensure the high quality and business relevance of the model input data, and provide a reliable guarantee for accurate evaluation and risk warning.

[0090] Milestone information data is the key node data used to track and evaluate the project progress in ICT project management, covering planned and actual completion times, delays, associated tasks, etc., and reflecting the stage achievements and risk points of project execution. Examples are as Figure 2 shown.

[0091] Furthermore, the method further includes:

[0092] Conduct exploratory data analysis on the obtained original data, including:

[0093] Conduct data checks, including: consistency with the data dictionary, information sufficiency, data understanding, data availability, data relevance;

[0094] Accounts receivable analysis: Analyze the scale of accounts receivable and the changing trend of the aging structure from the aspects of province, project type, and statistical month;

[0095] Based on the statistical report, view the increasing and decreasing trend and interval distribution of revenue;

[0096] Based on the statistical reports, view the distribution of gross profit margins and the changing trends over different years;

[0097] Based on the statistical reports, view the overall proportion of deferred revenue collection, the increasing and decreasing trends, and the distribution of days.

[0098] Through Exploratory Data Analysis (EDA), it is a crucial step to ensure data quality, understand business rules, and support model design. It can explore the data and initially grasp the situation, including:

[0099] 1) Conduct data checks to confirm the following: consistency with the data dictionary (such as table structure, etc.); sufficiency of information (such as fields, granularity, etc.); understanding of the data (such as the meaning of processed fields, statistical caliber, processing logic, etc.); usability (whether there are cases that require cleaning work such as deduplication, screening, format regularization, missing value filling, error and contradiction correction, etc.); correlation (further understand the relevance between data by analyzing the correlation between fields / variables).

[0100] 2) Accounts receivable analysis: Analyze the scale of accounts receivable and the changing trends of the aging structure from the aspects of provinces, project types, and statistical months.

[0101] 3) Based on the statistical reports, view the increasing and decreasing trends of revenue and the interval distribution.

[0102] 4) Based on the statistical reports, view the distribution of gross profit margins and the changing trends over different years.

[0103] 5) Based on the statistical reports, view the overall proportion of deferred revenue collection, the increasing and decreasing trends, and the distribution of days.

[0104] Through the above analysis, data quality diagnosis can be carried out: discover dirty data and outliers; check data integrity: identify missing fields (such as some projects missing milestone dates or financial data); Example: If it is found that 20% of the projects do not record the "actual implemented gross profit margin", the data needs to be completed or the model dimensions need to be adjusted; Detect outliers: Discover extreme values through box plots and scatter plots (such as negative or over 100% gross profit margins); Verify data consistency: Check whether the field definitions match the data dictionary (such as the "project status" field contains non-preset values); Example: If the "suspended" status appears in the "project status", its handling rules (whether to be included in the model) need to be clarified.

[0105] For the distribution analysis of a single indicator, the central tendency and dispersion degree of key indicators (such as gross profit margins, deferred days) can be analyzed by drawing histograms and density plots; For example: If it is found that 80% of the project gross profit margins are concentrated in the range of 5% - 15%, the scoring function for this interval needs to be optimized in the model.

[0106] For multi - indicator correlation analysis, the relationships between indicators can be explored through heatmaps and scatter matrices (such as whether there is a negative correlation between gross profit margin and the aging of accounts receivable); for example, projects with high gross profit margins usually have shorter aging periods, and collaborative risk warning rules can be designed based on this. For time - trend analysis, analyze the change trends of indicators by time dimension (such as quarters, years). For example, the number of delayed milestones soared in Q4 of 2023, and it is necessary to check whether it is due to external environment adjustments or insufficient resources.

[0107] Through exploratory data analysis (EDA), data reliability can be ensured (cleaning dirty data, handling outliers); business rules can be revealed (distribution, correlation, trend); model design can be guided (function form, threshold setting); management decisions can be empowered (risk positioning, resource optimization). Through EDA, operators can transform raw data into valuable intelligence, ensure that the health model not only conforms to data rules but also fits the business reality, and ultimately achieve refined project management.

[0108] Furthermore, the business scoring indicators include:

[0109] The project - established gross profit margin and the executed gross profit margin in the dimension of gross profit margin status;

[0110] The accounts - receivable - to - revenue ratio and the aging structure in the dimension of accounts receivable;

[0111] The current calculated days of delay, the number of milestones with delayed charges, the maximum historical days of delayed charges, and the historical number of milestones with delayed charges in the dimension of milestone fulfillment.

[0112] When constructing the scoring dimension, process information such as the gross profit margin status, accounts receivable, and milestone fulfillment of the project was comprehensively considered;

[0113] Gross profit margin status: In addition to the project - established gross profit margin, the executed gross profit margin indicator was added to reflect the change in gross profit margin during the actual execution of the project. The actual executed gross profit margin = (revenue - cost - advance payment - bad debts) / revenue * 100%, and the gross profit margin execution deviation rate = (actual executed gross profit margin - project - established gross profit margin) / project - established gross profit margin * 100%.

[0114] Milestone fulfillment: The current calculated days of delay, the number of milestones with delayed charges, the maximum historical days of delayed charges, and the historical number of milestones with delayed charges were constructed as indicators. Historical delayed charges consider the situation where the current charging of the project is completed, but there have been previous delayed charges, which more comprehensively evaluates the process risk of the project.

[0115] Accounts receivable status: The accounts - receivable - to - revenue ratio indicator and the aging structure indicator were constructed, taking into account both the scale and quality of accounts receivable.

[0116] When calculating the accounts receivable structure index, based on the account age rolling rate data and combined with the MA model (moving average model), the coefficients for different account ages are measured.

[0117] Furthermore, fitting the dynamic relationship between the business scoring index and the score through a non-linear function and optimizing the fitting parameters using the actual model data set includes:

[0118] Establish the upper and lower bounds of the index: Use the data within the observation period, analyze the distribution of each index of the data, and combine with the actual business situation to determine the upper and lower bounds of the index score change. The index scores exceeding the upper and lower bounds are given full marks or zero marks;

[0119] Prepare the data corresponding to the scores: Combine the index distribution related to the health degree and expert opinions to construct a data set;

[0120] Define the non-linear model: Based on the law of the health degree of the business changing with the index, determine the corresponding non-linear function and concavity / convexity, and define the MODEL function;

[0121] Fit the model: For each index, perform model fitting through the corresponding non-linear function;

[0122] Evaluate the model: Visualize the fitting effect of the non-linear scoring function by plotting the actual data points and the fitting curve;

[0123] Determine the parameters: Judge the fitting effect through the visualized data and the fitting curve, and determine the final parameter values;

[0124] Measurement of the account age structure coefficient: Establish the historical rolling rate data of the account age, and construct the MA model to measure the coefficients for different account ages.

[0125] During the fitting process of the scoring function, first, it is necessary to construct the scoring function, including:

[0126] 1. Establish the upper and lower bounds of the index;

[0127] Use the data within the observation period, analyze its index distribution, and combine with the actual business situation to determine the upper and lower bounds of its score change. The index scores exceeding the upper and lower bounds are given full marks or zero marks.

[0128] 2. Prepare the data corresponding to the scores;

[0129] Combine the index distribution related to the health degree and expert opinions to construct a data set. Taking the gross profit margin as an example, construct a set of data:

[0130] (GROSS_MARGIN_P_I, SCORE_I), where I takes values from [1, 10], to obtain the sequence data of X_DATA and Y_DATA.

[0131] The above data is used to construct a training dataset for the gross profit margin indicator to fit the health score function. By combining the historical data distribution and expert experience, a set of data pairs containing the gross profit margin value (X_DATA) and its corresponding health score (Y_DATA) is generated, providing a basis for subsequent model training.

[0132] By combining the historical data distribution and expert experience, construct a (gross profit margin, score) dataset to provide a training basis for the health assessment model, ensuring that the score function not only conforms to the data pattern but also is close to the business reality, ultimately achieving accurate risk warning and resource optimization.

[0133] 3. Define a non - linear model;

[0134] According to the law of the change of business health degree with indicators, determine the corresponding non - linear function and concavity - convexity, and define the MODEL function. Taking the gross profit margin as an example, SCORE = 30 * POWER(GROSS_MARGIN_P / 0.15, K), where K takes values in (0, 1).

[0135] 4. Fit the model;

[0136] Taking the gross profit margin as an example, use CURVE_FIT to fit the model.

[0137] INTIAL_PARAM = [0.001]

[0138] PARAM, COVARIANCE = CURVE_FIT(MODEL, X_DATA, Y_DATA, P0 = INITIAL_PARAM)

[0139] A = PARAMS

[0140] The above code uses the curve_fit method to fit the non - linear relationship between the gross profit margin (X_DATA) and the health score (Y_DATA), including:

[0141] Select a non - linear function that fits the business law;

[0142] Use curve_fit to optimize the parameters and improve the convergence efficiency by combining the initial guess.

[0143] 5. Evaluate the model;

[0144] X_FIT = NP.LINSPACE(0, 0.15, 10)

[0145] Y_FIT = MODEL(X_FIT, *PARAMS)

[0146] plt.plot(x_fit, y_fit, color='red', label='Fitted Curve')

[0147] plt.legend()

[0148] plt.title('NON-LINEAR FIT: INDEX MODEL')

[0149] plt.xlabel('X')

[0150] plt.ylabel('Y')

[0151] plt.show()

[0152] The above code is used to visualize the fitting effect of the non-linear scoring function. By plotting the actual data points and the fitted curve, it verifies the ability of the model to depict the business rules. The actual data points are obtained by pre-scoring each indicator in the model dataset according to the project status and expert experience.

[0153] Key Parameters and Business Meanings

[0154]

[0155] Model Verification: By comparing the distribution of the fitted curve and the original data points, it is judged whether the non-linear function reasonably depicts the change law of the indicators.

[0156] For example: when the gross profit margin increases from 5% to 10%, the score should increase significantly (the curve is steep), while the growth slows down from 10% to 15% (the curve is gentle).

[0157] Business Interpretability:

[0158] The curve shape in the figure reflects the business logic, such as:

[0159] Convex function: The change of the indicator at the initial stage has a great impact on the score, and the marginal utility decreases later (applicable to risk indicators).

[0160] Concave function: The indicator needs to reach a threshold before it has a significant impact on the score (applicable to profit indicators).

[0161] Basis for Parameter Tuning:

[0162] If the deviation between the fitted curve and the business expectation is large (for example, the score corresponding to a 10% gross profit margin is lower than the actual value), the initial parameters or the function form need to be adjusted.

[0163] When the gross profit margin is lower than 5%, the score drops sharply (high-risk area);

[0164] The gross profit margin of 5% - 10% is the key improvement interval;

[0165] When it exceeds 10%, the score growth slows down, and other dimensions (such as accounts receivable) need to be given priority attention.

[0166] The evaluation model is the core verification link in the health assessment model. Through visualization means, it ensures that the scoring function conforms to the business reality, providing a reliable basis for subsequent weight optimization and risk warning.

[0167] 6. Judge the fitting effect through visualized data and fitting curves, and determine the final parameter values;

[0168] 7. Measurement of the aging structure coefficient: Establish historical rolling rate data of the aging. The rolling rate of T aging period M1-3 -> M4-6 = the total amount of the T aging period that is in M1-M3 and in M4-M6 after 3 months / the total amount of the T aging period that is in M1-M3 * 100%. Then construct an MA model to measure the coefficients of different aging periods.

[0169] By analyzing the historical rolling rate of the accounts receivable aging, construct a dynamic coefficient model to quantify the impact of different aging periods on the project health, and solve the problem of "static weights cannot reflect the changes in aging risks" in traditional methods. MA model selection: Use Moving Average to smooth historical fluctuations and predict future rolling rates. The higher the rolling rate, the greater the risk of overdue accounts and the higher the deduction coefficient of the health score. Upgrade the static aging management to a dynamic risk prevention and control system, significantly improving the control efficiency of the accounts receivable of the operator's ICT projects.

[0170] Function and coefficient measurement examples are as Figure 3 shown.

[0171] In the health assessment model of the operator's ICT project, the fitting of the scoring function is the core step to convert business indicators (such as gross profit margin, delay days, accounts receivable, etc.) into standardized health scores. Its specific functions are as follows:

[0172] 1. Accurately map the non-linear relationship between business indicators and health;

[0173] Traditional linear or interval scoring methods assume a fixed proportional relationship between indicators and scores (e.g., "for every 10-day increase in the number of days of delay, 5 points are deducted"), which have significant drawbacks. For example, in the interval scoring method, if the gross profit margin is in the range of [0, 1%), 5 points are awarded, and in the range of [1%, 2%), 10 points are awarded. This treatment method results in insufficient precision, as a gross profit margin of 0 and 0.8% receive the same score. In the linear scoring method, such as SCORE = 3.5 - 0.005 * (HIS_DELAY_DAYS - 60), the scores decrease uniformly, but in reality, fewer points should be deducted when the delay is close to 60 days, and the number of points deducted increases faster as the number of days of delay increases. The linear scoring method cannot reflect the actual business rules. In actual business, the impact of indicator changes on health may be non-linear. Example: Number of days of delay: The impact of initial delay on the score is relatively small, but it increases sharply (exponential growth) after exceeding the threshold.

[0174] Gross profit margin: When it is lower than the target value, the score drops rapidly, and the marginal utility decreases after reaching the target (logarithmic or power function relationship).

[0175] The solution of the embodiments of the present disclosure fits the dynamic relationship between indicators and scores through a non-linear function (such as Score = 30 * (gross_margin_p / 0.15)^k), and optimizes the parameters (such as the k value) using actual data to make the score more in line with the business rules. The effect can be realized that when the gross profit margin increases from 5% to 10%, the score may increase by 30%; when it increases from 10% to 15%, it only increases by 10%, reflecting the diminishing marginal return.

[0176] 2. Improve the adaptability of the model to complex business scenarios:

[0177] Implement dynamic parameter optimization: Use tools such as curve_fit to fit the parameters. The model can automatically adjust the function form according to different project types (such as 5G construction, smart city) to meet diverse business needs. For example, for high-risk projects: the coefficient of the aging rolling rate may adopt a steeper curve (strictly control the overdue risk); for long-cycle projects: the deduction rule for milestone delays may be more lenient (tolerate reasonable fluctuations).

[0178] The fitting process depends on actual historical data to ensure that the scoring rules are based on real business performance rather than subjective assumptions.

[0179] 3. Support the collaborative analysis and weight allocation of multi-dimensional indicators:

[0180] Unify the scoring scale:

[0181] Convert indicators with different dimensions (such as percentages, days, amounts) into unified health scores, which is convenient for subsequent weighted aggregation.

[0182] Example:

[0183] Gross profit margin (0% - 15%) → 0 - 30 points;

[0184] Number of delay days (0 - 60 days) → 0 - 20 points.

[0185] The scoring function provides standardized input for the Analytic Hierarchy Process (AHP) and Lagrangian function optimization, ensuring that the weight allocation reflects both expert experience and data patterns.

[0186] 4. Enhance the transparency and interpretability of the model:

[0187] Visual verification:

[0188] By plotting the fitting curve (such as plt.plot(x_fit, y_fit)), visually display the relationship between the indicators and the scores, facilitating business personnel to understand the model logic.

[0189] Parameter traceability:

[0190] The fitted parameters (such as k = 0.5) can be archived and compared with historical data to support model iteration and auditing.

[0191] The scoring function fitting converts the original business data into a health score with business significance, solving the pain points of traditional methods such as "single dimension, poor adaptability, and weak interpretability". Nonlinear mapping can accurately depict business rules; parameter optimization can dynamically adapt to complex scenarios; through this step, operators can achieve the transformation from "experience-driven" to "data-driven", significantly improving the refinement level of ICT project management.

[0192] Furthermore, the determination of the sub-item weights of each business scoring indicator includes:

[0193] Generate a suitable weight range for each indicator based on expert evaluation and the AHP analytic hierarchy process, including:

[0194] Construct a judgment matrix by pairwise comparing the importance of each indicator by experts;

[0195] Calculate the weights given by each expert through the geometric mean method and perform normalization processing. Based on the mean and standard deviation of the weights, and combined with relevant rule-based business adjustments, set the corresponding weight constraint intervals;

[0196] Based on the Lagrangian multiplier method, perform weight optimization to find the optimal ratio of the weights of the multi-dimensional scoring system, including:

[0197] Construct a project scoring objective function based on multiple indicators, and at the same time define the normalization constraints and interval constraints for each weight;

[0198] Construct the Lagrangian function and introduce the Lagrange multiplier system. Through the strict mathematical framework of the KKT conditions, model and solve the multi-constraint optimization problem to ensure that the optimization result not only meets the extreme value requirements of the objective function but also conforms to the preset weight constraint conditions.

[0199] In terms of the sub-item weight setting, in the embodiments of the present disclosure, the appropriate weight interval for each index is generated according to expert evaluation and the analytic hierarchy process (AHP). Then, based on the Lagrange multiplier method, the weight is optimized to find the optimal ratio of the weights of the multi-dimensional scoring system, making the weight setting more in line with the business reality.

[0200] 1. AHP analytic hierarchy process and expert evaluation

[0201] First, the experts make pairwise comparisons of the importance of each index, using the 1-9 scale method. For example, if the experts believe that the project gross profit margin is significantly more important than the execution gross profit margin, then mark 5 in the corresponding position in the judgment matrix; conversely, mark 1 / 5 for the importance of the execution gross profit margin relative to the project gross profit margin. Taking these 8 indexes as an example, construct the judgment matrix:

[0202] A=(a ij ),

[0203] where i, j = 1, 2, 3,... 8, a ij represents the importance scale of the i-th index relative to the j-th index;

[0204] Calculate the weights given by each expert through the geometric mean method and perform normalization processing. Based on its mean and standard deviation, and combined with relevant rules for business adjustment, set the corresponding weight constraint interval, as shown in Figure 4 shown.

[0205] 2. Construct the project scoring objective function based on multiple indexes, and at the same time define the normalization constraint and interval constraint of each weight

[0206] The project scoring formula is:

[0207]

[0208] where, represents the predicted health score of the i-th project, ω j represents the weight of the j-th index, f j (x i,j ) represents the non-linear scoring function of the j-th index

[0209] Ensure that the sum of the weights is 1 through the normalization constraint, and the calculation formula is as follows:

[0210]

[0211] Interval constraint:

[0212]

[0213] Among them, represents the lower bound of the interval, represents the upper bound of the interval.

[0214] 3. Construct the Lagrangian function and introduce the Lagrange multiplier system, and achieve precise modeling and solution of the multi-constraint optimization problem through the strict mathematical framework of the KKT conditions. Through the synergistic effect of the gradient balance condition, the complementary slackness criterion, and the non-negativity constraint of the multiplier, a saddle-point equilibrium state is constructed within the feasible solution domain to ensure that the optimization result not only meets the extreme value requirements of the objective function but also conforms to the preset weight constraint conditions.

[0215] The calculation formula of the Lagrangian function is as follows:

[0216]

[0217] Among them, λ represents the Lagrange multiplier, and μ j , v j represent the Lagrange multipliers;

[0218] The function L is partially differentiated with respect to λ, μ j , v j , w j to obtain a system of equations and solve it:

[0219]

[0220] 4. Through the Analytic Hierarchy Process (AHP) and expert evaluation, and with the help of the precise derivation and solution of the Lagrange equations, considering the mutual relationships and constraint conditions between different dimensions and indicators, and supported by the optimization theory, the accurate optimization of the weights of each dimension and indicator is achieved under the condition of meeting various constraint conditions. The weight distribution results are as Figure 5 shown.

[0221] Through expert evaluation and AHP to generate the initial weight interval + Lagrangian function optimization, the operator ICT project health model has achieved: the balance between scientificity and flexibility: respecting expert experience while relying on data-driven optimization. Precise risk control: High-weight indicators (such as gross profit margin) directly affect the score, promoting resource focus on key issues. Dynamic adjustment ability: Update the weight constraints with business changes to ensure the long-term effectiveness of the model.

[0222] Furthermore, the method further includes:

[0223] Determine the health level of the project according to the measured result of the project health score with reference to the project score grading standard;

[0224] Warn about projects with a significant decline in scores or projects that do not meet the health level standard.

[0225] The grading criteria for project scores are as Figure 6 shown,

[0226] Based on the calculated score of the project health score, the health level of the project can be obtained, and different treatments can be carried out according to its level. And warn about projects with a significant decline in scores or projects that do not meet the health level standard. For example, through real-time scoring calculation by the model, a health score can be generated based on the latest project data, triggering a warning (such as marking as "dangerous" when the score < 70); for a certain project, due to the accounts receivable aging exceeding 90 days and the aging structure coefficient dropping sharply → the health score drops from 85 points to 65 points → the system automatically notifies the risk control team to intervene. By analyzing low-scoring items (such as too high gross profit margin deviation rate), locate the crux of the project and allocate resources accordingly (such as increasing the budget and adjusting the milestone plan).

[0227] The overall scheme process of the embodiments of the present disclosure is as Figure 7 , including:

[0228] Data layer

[0229] Subscribe / Obtain data: Integrate multi-source data (project information, finance, milestones, etc.).

[0230] Data preprocessing: Clean outliers, standardize the format, and generate derived indicators.

[0231] Model layer

[0232] Determine scoring dimensions: Define evaluation dimensions such as finance, progress, and risk.

[0233] Fit the scoring function: Construct a non-linear function to quantify the dynamic relationship between indicators and health.

[0234] Optimize weight allocation: Combine AHP (expert experience) and the Lagrange multiplier method (data-driven) to determine the optimal weights.

[0235] Application layer

[0236] Measure and rate health: Calculate the comprehensive score and classify (green / yellow / red), triggering a warning.

[0237] Output and application: Empower management decisions (such as resource allocation and customer communication) through visual dashboards and warning notifications.

[0238] The project health assessment method of the embodiments of the present disclosure takes the original input as the project basic information, milestone information, financial information such as revenue and cost, and data related to the accounts receivable aging; the final output includes:

[0239] 1) Project health model, which is used to measure the current health of a project.

[0240] 2) Dynamic monitoring and early warning function. For projects with a significant decline in scores or major projects with potential problems, the system can issue warnings in a timely manner to remind relevant personnel to take necessary measures to achieve early prevention and rapid response.

[0241] 3) Visualization display tool. A variety of types of dashboards (such as large-screen display, ICT view management module) are developed to intuitively present the health levels of projects under the jurisdiction of each province, facilitating the management layer to effectively manage the overall risk.

[0242] The sample score chart of some sub-projects in a certain province obtained by the above method is as Figure 8 shown. Applying the model scoring results to the early warning platform and ICT management view can guide business personnel to take corresponding treatment measures for projects with low health levels, and at the same time assist the management layer to control the overall project risk in real time.

[0243] In the embodiments of the present disclosure, an ICT project health evaluation model is constructed from dimensions such as project gross profit margin, milestone fulfillment, and accounts receivable, realizing a full-range scan of the health of ICT projects. The model uses the non-linear function method, adopts the optimal order, more accurately depicts the impact of index changes on the health level, and organically integrates expert scoring, the analytic hierarchy process (AHP), and the Lagrange function optimization method, and combines with the MA model to determine the weights of the scoring indicators for each dimension; making the scoring model more adaptable to business changes. By scoring each ICT project, the current health status of the project is accurately judged. The higher the score, the better its health level, and its health level is divided according to the score range. It provides core empowerment for management scenarios such as project risk control and accounts receivable control, and improves the scientific and refined level of project management.

[0244] Embodiment 2 of the present disclosure also provides a project health evaluation system, as Figure 9 shown: The system includes:

[0245] A data selection module 11, which is set to select projects to be monitored according to the actual business scenario and construct a model data set;

[0246] A scoring index construction module 12, which is set to comprehensively consider the gross profit margin situation, accounts receivable, and milestone fulfillment information of the project, and construct each sub-item business scoring indicators corresponding to the gross profit margin dimension, accounts receivable dimension, and milestone fulfillment dimension;

[0247] A scoring function fitting module 13, which is set to fit the dynamic relationship between the business scoring indicators and the score through a non-linear function, and optimize the fitting parameters using the actual model data set to convert the business scoring indicators into standardized health scores, making the scores more in line with business rules;

[0248] Sub - item weight determination module 14, which is set to determine the sub - item weights of each business scoring indicator;

[0249] Healthiness measurement module 15, which is set to calculate the healthiness score of the project by using the latest data as of the current accounting period and based on the scoring function and the determined sub - item weights.

[0250] Furthermore, the data selection module 11 is specifically set as follows:

[0251] When selecting the projects to be monitored, eliminate the projects whose establishment time exceeds the preset time and the projects in abnormal status;

[0252] Obtain the original data of the projects from the data middle - platform in the form of subscribing to data sets;

[0253] Obtain the project basic information, milestone information, financial information and information related to the accounts receivable aging of each project from the original data.

[0254] Furthermore, the system further includes a data exploration module 16;

[0255] The data exploration module 16 is set to perform exploratory data analysis on the obtained original data, including:

[0256] Perform data checks, including: consistency with the data dictionary, sufficiency of information, data understanding, data availability, data relevance;

[0257] Accounts receivable analysis: Analyze the scale of accounts receivable, the changing trend of the aging structure from the aspects of province, project type, and statistical month;

[0258] Based on the statistical reports, view the increasing and decreasing trend and the interval distribution of revenues;

[0259] Based on the statistical reports, view the distribution of gross profit margins and the changing trend in different years;

[0260] Based on the statistical reports, view the overall proportion of delayed collection, the increasing and decreasing trend, and the distribution of days.

[0261] Furthermore, the business scoring indicators include:

[0262] The established gross profit margin and the executed gross profit margin in the dimension of gross profit margin status;

[0263] The accounts receivable - to - revenue ratio and the aging structure in the dimension of accounts receivable;

[0264] The current days of delayed calculation, the number of milestones of delayed collection, the maximum value of historical days of delayed collection, and the number of historical milestones of delayed collection in the dimension of milestone fulfillment.

[0265] Further, the scoring function fitting module 13 is specifically configured as follows:

[0266] Establish the upper and lower bounds of the indicators: Use the data within the observation period to analyze the distribution of each indicator of the data. Combine the actual business situation to determine the upper and lower bounds of the indicator score changes. The indicator values exceeding the upper and lower bounds are scored as full marks or zero points.

[0267] Prepare the data corresponding to the scores: Combine the indicator distribution related to the health level and expert opinions to construct a data set.

[0268] Define the non-linear model: Based on the law of the health level changing with the indicators in the business, determine the corresponding non-linear function and concavity / convexity, and define the MODEL function.

[0269] Fit the model: For each indicator, perform model fitting through the corresponding non-linear function.

[0270] Evaluate the model: Visualize the fitting effect of the non-linear scoring function by plotting the actual data points and the fitting curve.

[0271] Determine the parameters: Judge the fitting effect through the visualized data and the fitting curve, and determine the final parameter values.

[0272] Calculate the aging structure coefficient: Establish the historical rolling rate data of the aging, and construct an MA model to calculate the coefficients for different aging periods.

[0273] Further, the sub-item weight determination module 14 is specifically configured as follows:

[0274] Generate a suitable weight interval for each indicator according to expert evaluation and the AHP (Analytic Hierarchy Process), including:

[0275] Construct a judgment matrix by pairwise comparing the importance of each indicator by experts.

[0276] Calculate the weights given by each expert through the geometric mean method and perform normalization processing. Based on the mean and standard deviation of the weights, and combined with relevant rule-based business adjustments, set the corresponding weight constraint intervals.

[0277] Perform weight optimization based on the Lagrange multiplier method to find the optimal ratio of the weights of the multi-dimensional scoring system, including:

[0278] Construct a project scoring objective function based on multiple indicators, and at the same time define the normalization constraint and interval constraint of each weight.

[0279] Construct a Lagrangian function and introduce the Lagrange multiplier system. Through the strict mathematical framework of the KKT (Karush-Kuhn-Tucker) conditions, realize the modeling and solution of the multi-constraint optimization problem, ensuring that the optimization result not only meets the extreme value requirements of the objective function but also conforms to the preset weight constraint conditions.

[0280] Furthermore, the healthiness measurement module 15 is further configured to:

[0281] Determine the healthiness level of the project according to the healthiness score measurement result of the project with reference to the project score grading standard;

[0282] Issue a warning for projects with a significant decline in scores or projects whose healthiness levels do not meet the standards.

[0283] The project healthiness evaluation system of the embodiments of the present disclosure is used to implement the project healthiness evaluation method in the first method embodiment, so the description is relatively simple. For specific details, reference can be made to the relevant descriptions in the previous method embodiments, which will not be elaborated here.

[0284] In addition, as Figure 10 shown, the third embodiment of the present disclosure further provides an electronic device, including a memory 100 and a processor 200. A computer program is stored in the memory 100. When the processor 200 runs the computer program stored in the memory 100, the processor 200 executes the above various possible methods.

[0285] Among them, the memory 100 is connected to the processor 200. The memory 100 can adopt flash memory, read-only memory or other memories, and the processor 200 can adopt a central processing unit or a single-chip microcomputer.

[0286] In addition, the embodiments of the present disclosure further provide a computer-readable storage medium, on which a computer program is stored, and the computer program is executed by the processor to implement the above various possible methods.

[0287] The computer-readable storage medium includes volatile or non-volatile, removable or non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, computer program modules or other data). The computer-readable storage medium includes, but is not limited to, RAM (Random Access Memory), ROM (Read-Only Memory), EEPROM (Electrically Erasable Programmable Read Only Memory), flash memory or other memory technologies, CD-ROM (Compact Disc Read-Only Memory), digital versatile disc (DVD, Digital Video Disc) or other optical disc storage, magnetic cassette, tape, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer.

[0288] It is understandable that the above embodiments are merely exemplary embodiments adopted to illustrate the principles of the present disclosure. However, the present disclosure is not limited thereto. For those of ordinary skill in the art, various modifications and improvements can be made without departing from the spirit and essence of the present disclosure, and these modifications and improvements are also regarded as the protection scope of the present disclosure.

Claims

1. A project health assessment method, characterized in that, The method includes: Select the items to be monitored according to the actual business scenario and construct a model data set; Comprehensively consider the gross profit margin status, accounts receivable, and milestone fulfillment information of the items, and construct various sub-item business scoring indicators corresponding to the gross profit margin status dimension, accounts receivable dimension, and milestone fulfillment dimension respectively; Fit the dynamic relationship between the business scoring indicators and the scores through a non-linear function, and optimize the fitting parameters using the actual model data set to convert the business scoring indicators into standardized health scores, making the scores more in line with business rules; Determine the sub-item weights of each business scoring indicator; Adopt the latest data as of the current accounting period, and calculate the project health score according to the scoring function and the determined sub-item weights.

2. The method according to claim 1, wherein The step of selecting the items to be monitored according to the actual business scenario and constructing a model data set includes: When selecting the items to be monitored, exclude the items whose project establishment exceeds the preset time and the items in abnormal status; Obtain the original data of the projects in the form of subscribing to the data set from the data middle platform; Obtain the project basic information, milestone information, financial information, and accounts receivable aging-related information of each project from the original data.

3. The method according to claim 2, wherein The method further includes: Conduct exploratory data analysis on the obtained original data, including: Conduct data checks, including: consistency with the data dictionary, information sufficiency, data understanding, data availability, and data relevance; Accounts receivable analysis: Analyze the scale of accounts receivable, the changing trend of the aging structure from the aspects of province, project type, and statistical month; Based on the statistical report, view the increasing and decreasing trend and interval distribution of revenue; Based on the statistical report, view the distribution of the gross profit margin and its changing trend in different years; Based on the statistical report, view the overall proportion of delayed collection, the increasing and decreasing trend, and the number of days distribution.

4. The method according to claim 1, characterized in that, The business scoring indicators include: The established gross profit margin and the executed gross profit margin in the gross profit margin status dimension; The accounts receivable ratio to revenue and the aging structure in the accounts receivable dimension; The current number of days of delayed calculation, the number of milestones with delayed collection, the maximum value of the historical number of days of delayed collection, and the historical number of milestones with delayed collection in the milestone fulfillment dimension.

5. The method according to claim 3, characterized in that The step of fitting the dynamic relationship between the business scoring indicators and the scores through a non-linear function and optimizing the fitting parameters using the actual model data set includes: Establish the upper and lower bounds of the indicators: Use the data within the observation period, analyze the distribution of each indicator of the data, and combine the actual business situation to determine the upper and lower bounds of the change of the indicator scores. The indicator scores exceeding the upper and lower bounds are full marks or zero marks; Prepare the data corresponding to the scores: Combine the distribution of the indicators related to health and the opinions of experts to construct a data set; Define the non-linear model: According to the law of the change of business health degree with indicators, determine the corresponding non-linear function and concavity and convexity, and define the MODEL function; Fit the model: For each indicator, perform model fitting through the corresponding non-linear function; Evaluate the model: Visualize the fitting effect of the non-linear scoring function by plotting the actual data points and the fitting curve; Determine the parameters: Judge the fitting effect through the visualized data and the fitting curve, and determine the final parameter values; Calculate the aging structure coefficient: Establish the historical rolling rate data of the aging, and construct an MA model to calculate the coefficients of different aging periods.

6. The method according to claim 1, wherein The determination of the sub-item weights of each business scoring indicator includes: Generating a suitable weight range for each indicator based on expert evaluation and the Analytic Hierarchy Process (AHP), including: Constructing a judgment matrix by pairwise comparison of the importance of each indicator by experts; Calculating the weights given by each expert through the geometric mean method and normalizing them, setting corresponding weight constraint intervals based on the mean and standard deviation of the weights and combining relevant rules for business adjustment; Optimizing the weights based on the Lagrange multiplier method to find the optimal ratio of the weights of the multi-dimensional scoring system, including: Constructing a project scoring objective function based on multiple indicators, and at the same time defining the normalization constraint and interval constraint of each weight; Constructing a Lagrangian function and introducing a Lagrange multiplier system, and realizing the modeling and solution of the multi-constraint optimization problem through the strict mathematical framework of the KKT conditions, ensuring that the optimization result not only meets the extreme value requirements of the objective function but also conforms to the preset weight constraint conditions.

7. The method according to claim 1, wherein The method further includes: Determining the health level of the project according to the measurement result of the project's health score with reference to the project score grading standard; Issuing a warning for projects with a significant decline in scores or projects that do not meet the health level standard.

8. A project health assessment system, characterized in that, The system includes: A data selection module, which is set to select the projects to be monitored according to the actual business scenario and construct a model data set; A scoring indicator construction module, which is set to comprehensively consider the gross profit margin status, accounts receivable, and milestone fulfillment information of the project, and construct each sub-item business scoring indicator corresponding to the gross profit margin status dimension, accounts receivable dimension, and milestone fulfillment dimension; A scoring function fitting module, which is set to fit the dynamic relationship between the business scoring indicator and the score through a non-linear function, and optimize the fitting parameters using the actual model data set to convert the business scoring indicator into a standardized health score, making the score more in line with business rules; A sub-item weight determination module, which is set to determine the sub-item weights of each business scoring indicator; A health measurement module, which is set to measure the health score of the project using the latest data as of the current accounting period, based on the scoring function and the determined sub-item weights.

9. An electronic device, characterized in that, It includes a memory and a processor. A computer program is stored in the memory. When the processor runs the computer program stored in the memory, the processor executes the project health assessment method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium. When the computer program is executed by the processor, it realizes the project health assessment method according to any one of claims 1-7.

Images